Testing Clinical Intuitions About Barriers to Improvement in Cognitive-Behavioral Therapy for Panic Disorder. Issue 4 (July 2021)

• Record Type: Journal Article
• Authors: Schwartz, Rachel A.; Chambless, Dianne L.; Barber, Jacques P.; Milrod, Barbara
• Is Part Of: Behavior therapy. Volume 52:Issue 4(2021)
• Page Start: 956
• Page End: 969
• Publication Date: 2021-07
• Subjects: clinical intuition -- panic disorder -- agoraphobia -- psychotherapy outcome prediction -- psychotherapy moderation; Behavior therapy -- Periodicals; 616.8914205
• DOI: 10.1016/j.beth.2020.12.004
• Languages: English
• ISSNs: 0005-7894

Abstract

Highlights: No clinical intuitions were supported in this clinical trial. Dissociating during panic was related to better outcome in CBT and dynamic therapy. CBT outperformed dynamic therapy for panic patients with comorbid PTSD. CBT outperformed dynamic therapy for patients with less severe panic. Research can utilize input from clinicians to help close the practice-research gap. Abstract: Although clinical intuitions influence psychotherapeutic practice and are a rich source of novel hypotheses for research, many remain to be empirically tested. This study evaluates whether clinicians' beliefs about barriers to progress in cognitive-behavioral therapy (CBT) for panic disorder are supported by data. Data from a randomized-controlled trial comparing CBT to panic-focused psychodynamic psychotherapy (PFPP) for adults with primary panic disorder ( N = 161) were used to evaluate 15 factors endorsed by clinicians as impediments to CBT in a recent survey. Panic severity was assessed before, during (at Weeks 1, 5, and 9), and at termination of treatment (Week 12) using the Panic Disorder Severity Scale. Hierarchical linear modeling revealed that none of the perceived barriers were predictive of poor outcome. Contrary to clinicians' intuitions, dissociation during panic attacks was associated with greater symptomatic improvement in both treatment arms ( β = −0.69, p < .05), above the effect of established predictors. Moderation analyses revealed that when patients had PTSD diagnosed with the Anxiety Disorders Interview Schedule ( β = 1.71, p < .05) or less severe panic disorder ( β = 0.45, p = .04), they changed more rapidly in CBT than in PFPP. Overall, clinician agreement was inversely related to the strength of a predictor ( r = −.24, p = .39). Although clinical intuitions can be useful as clinical and empirical signals, such beliefs should be critically examined before informing practice. Dialogue between academics and clinicians might be enhanced through research that incorporates input from front-line practitioners.